ADB Sustainable Development Working Paper Series. Approaches to Combat Hunger in Asia and the Pacific

Transcription

1 ADB Sustainable Development Working Paper Series Approaches to Combat Hunger in Asia and the Pacific Shiladitya Chatterjee, Amitava Mukherjee, and Raghbendra Jha No. 11 August 2010

2 ADB Sustainable Development Working Paper Series Approaches to Combat Hunger in Asia and the Pacific Shiladitya Chatterjee, Amitava Mukherjee, and Raghbendra Jha No. 11 August 2010 Shiladitya Chatterjee is an advisor at ADB's Strategy and Policy Department. Amitava Mukherjee is senior expert, Macroeconomic Policy and Development Division, United Nations Economic and Social Commission for Asia and the Pacific, and head, United Nations Centre for Alleviation of Poverty through Secondary Crops for Asia and the Pacific. Raghbendra Jha is a professor and director of the South Asia Center, Australian National University.

3 Asian Development Bank 6 ADB Avenue, Mandaluyong City 1550 Metro Manila, Philippines by Asian Development Bank August 2010 ISSN Publication Stock No. WPS The views expressed in this publication are those of the authors and do not necessarily reflect the views and policies of ADB, its Board of Governors, or the governments they represent. ADB does not guarantee the accuracy of the data included in this publication and accepts no responsibility for any consequence of their use. By making any designation of or reference to a particular territory or geographic area, or by using the term "country" in this document, ADB does not intend to make any judgments as to the legal or other status of any territory or area. This working paper series is maintained by the Regional and Sustainable Development Department. Other ADB working paper series are on economics, regional cooperation, and ADBI Working Paper Series. Further ADB publications can be found at ww.adb.org/publications/. The purpose of the series is to disseminate the findings of work in progress to encourage the exchange of ideas. The emphasis is on getting findings out quickly even if the presentation of the work is less than fully polished.

4 Contents Abbreviations... iii Acknowledgments... v 1. Introduction Hunger Definitions Estimates and Trends of Hunger and Poverty in the Region...3 Trends in Poverty and Hunger...3 Location of Hunger Hotspots by Subregion and Country...5 The Global Food and Economic Crises and Impact on Poverty and Hunger...7 Future Projections Child Hunger Causes and Correlates of Hunger...11 Poverty and Hunger...11 Economic Growth...13 Looking beyond Economic Growth for Factors Affecting Hunger...16 Literacy...18 Availability and Affordability of Food...19 Gender and Other Institutional Factors...24 Vulnerabilities, Risks, and Shocks...26 Hidden Hunger: Micronutrients and Clean Water Conclusions: Major Elements of a Hunger Strategy...29 References...34

5 Abbreviations ADB Asian Development Bank ASEAN Association of Southeast Asian Nations CIDA Canadian International Development Agency FAO Food and Agriculture Organization of the United Nations GDP gross domestic product GHI global hunger index Lao PDR Lao People's Democratic Republic MDG Millennium Development Goal PPP purchasing power parity PRC People's Republic of China UNDP United Nations Development Programme UNESCAP United Nations Economic and Social Commission for Asia and the Pacific UNICEF United Nations Children's Fund WHO World Health Organization Note: In this document, "$" refers to US dollars.

6 Acknowledgments This paper is a contribution to the Asian Development Bank (ADB) United Nations Economic and Social Commission for Asia and the Pacific (UNESCAP) United Nations Development Programme (UNDP) technical paper series on the Millennium Development Goals. It has benefited from a wide range of comments and inputs. The authors would like to especially thank the Partnership for Ending Child Hunger and Undernutrition, REACH, in which ADB was a partner and from which several conclusions on child hunger have been drawn; Thangavel Palanivel of UNDP; and Guanghua Wan of ADB for invaluable comments and suggestions. The authors would also like to acknowledge the contribution of Romana Mary Margaret Domingo, who provided data support and helped with the figures, charts, regressions, and final editing of this paper.

7

8 Approaches to Combat Hunger in Asia and the Pacific 1 1. Introduction Asia and the Pacific, despite visible signs of prosperity due to years of rapid economic growth, has made insufficient progress in freeing its population from hunger and malnutrition, the most basic marker of true development. Given the importance of reducing poverty and hunger as core development priorities, the international community has placed these as the first Millennium Development Goal (MDG), adopted in the historic Millennium Declaration of September The MDGs envision halving the incidence of poverty and hunger by 2015 and call for major improvements in the provision of basic services, promising a better life to millions of poor across the globe. However, if progress in reducing poverty and hunger is not accelerated in the region, such hopes will remain unrealized. The extreme poverty in Asia and the Pacific that affects around 900 million people, as well as the persistence of hunger in well over half a billion, questions the long-term sustainability of the region s economic growth and development. About one in every six persons in the region suffers from malnourishment, and about one in three children is underweight. These problems are particularly severe in South Asia and Southeast Asia. The MDG target to reduce hunger by half is unlikely to be achieved in the case of child hunger; the region may also miss the target for undernourishment, particularly if restored economic growth from the current slowdown is sluggish and if food prices start increasing. The task before the region's governments to achieve this crucial development goal is thus demanding, requiring greater prioritization of efforts and the adoption of new, more efficient approaches. Further, more regional and international efforts are also needed. The primary purposes of this paper are to trace the progress of efforts to (i) reduce hunger in Asia and the Pacific, (ii) identify reasons for their successes and failures, and (iii) suggest policy initiatives to help make tangible progress on these first MDG targets in the time remaining to Hunger Definitions There is no single, universally accepted definition of hunger or of its measurement. As hunger has become better understood over the years, however, the discussion has shifted from the mere availability of food to its access, affordability, and nutrition. In its report on the World Food Summit of 1996, the Food and Agriculture Organization (FAO) posited that hunger can be eliminated "when all people, at all times, have physical and economic access to sufficient, safe and nutritious food to meet their dietary needs and food preferences for an active and healthy life (FAO 1996). MDG goal 1 focuses on reducing the incidence of extreme poverty and hunger deprivation, and has two targets (United Nations Development Programme [UNDP] 2008). While the first target envisages halving of the proportion of poor between 1990 and 2015, the second target aims to halve the proportion of people who suffer from hunger by 2015, with the following indicators to measure progress: (i) (ii) prevalence of underweight children under age 5 years (United Nations Children's Fund [UNICEF] definition), and proportion of population below the minimum level of dietary energy consumption (FAO definition).

9 2 ADB Sustainable Development Working Paper Series No. 11 These hunger indicators provide two critical dimensions: child malnutrition measured by low weight and lack of sufficient dietary energy consumption. While other indicators exist, these two are most often referred to, partly owing to the availability of data. However, while experts recommend a minimum of 2,200 calories for adults undertaking light activity, some suggest other standards for measuring hunger. 1 The MDG target of undernourishment in the general population uses the FAO definition, which derives a proportion of the undernourished through distribution of the dietary energy supply within a country's population and by establishing a minimum cut-off point, below which persons are considered energy-consumption deficient (Box 2). The energy-deficiency measure of hunger has additional drawbacks, including failing to include other vital nutrients, such as minerals and vitamins; 2 different energy requirements of men, women, and children; and varying needs of persons engaged in different types of work. As seen in Appendix 1, the minimum daily nutritional requirements in rural India, recommended by an expert group on nutrition, vary greatly for age, gender, and occupational status. As a single measure of hunger, the total population suffering from energy-consumption deficiency has limitations. Thus, many development agencies have advocated also focusing on the most vulnerable group in the population children. Examining child malnutrition is also critical due to the large costs that the problem causes for labor and overall economic productivity. Thus, the second MDG indicator on overall hunger that is, child hunger came about. Attempts to construct composite indexes using both hunger indicators have also been made. The global hunger index (GHI), developed by the International Food Policy Research Institute (IFPRI), for example, combines three equally weighted indicators: (i) the proportion of people who are food-energy deficient, as estimated by FAO; (ii) the prevalence of underweight in children under age 5 years, as estimated by the World Health Organization (WHO); and (iii) the under age 5 years mortality rate, estimated by UNICEF. This measure suffers from the deficiencies associated with composite measures. All three measures, for example, are arbitrarily weighted equally, and the quality of the data used varies widely across countries, which aggregation may compound (Ahmed et al. 2007a). While improvements in hunger measures are necessary, the rest of this paper uses the two main MDG hunger indicators. 1 For example, Ahmed et al. (2007a and 2007b) defined three categories of the hungry according to the number of calories consumed per day: (i) subjacent hungry, consuming 1,800 2,200 calories per day; (ii) medial hungry, consuming 1,600 1,800 calories per day; and (iii) ultra hungry, consuming less than 1,600 calories per day. Regarding its definition of hunger, the International Food Policy Research Institute uses the popular $1 per day per person (adjusted for purchasing power parity [PPP]) as the cut-off point for poverty. The poor are further classified according to intensity of poverty being experienced: (i) subjacent poor, living on $0.75 $1.00 per day; (ii) medial poor, living on $0.50 $0.75 per day; and (iii) ultra poor, living on less than $0.50 a day. The authors find that there is strong association only between ultra poverty and hunger. 2 There is further discussion of micronutrients in section 5. For an analysis of micronutrient deprivation and its relation to calorie deprivation in rural India, see Jha et al. (2009a).

10 Approaches to Combat Hunger in Asia and the Pacific 3 3. Estimates and Trends of Hunger and Poverty in the Region This section studies (i) how the incidence of hunger has reduced over time, particularly in comparison to poverty; (ii) which countries have the most hungry by MDG hunger definitions; (iii) how the current economic and food crises have impacted hunger; and (iv) whether the MDG hunger targets are likely to be achieved. 3.1 Trends in Poverty and Hunger Asia and the Pacific witnessed a rapid reduction in poverty between 1990 and 2005 (Tables 1 and 2). However, the incidence of hunger and the number of hungry have not fallen as dramatically, indicating that hunger is a more stubborn adversary than poverty, necessitating different instruments for engagement. Table 1: Incidence of Poverty and Hunger (%, $1.00 PPP per day) Headcount ratio (%, $1.25 PPP per day) Prevalence of undernourishment (% of population) b Region or subregion Proportion of underweight children under age 5 years c,d Earliest Latest ( ) ( ) Asia and the Pacific East Asia Southeast Asia South Asia Central and West Asia Pacific a PPP = purchasing power parity. a The Pacific includes Papua New Guinea only, owing to scant data availability. b For prevalence and number of undernourished persons, East Asia s regional average includes the Democratic People's Republic of Korea and the Republic of Korea, while Southeast Asia s regional average includes Myanmar. c Country weights for proportion and number of underweight children under age 5 years were calculated by the Asian Development Bank (ADB) staff using 2005 population of children age 0 4 years from the 2008 Revision of the United Nations World Population Prospects. d For proportion and number of underweight children under age 5 years, the range of years for the earliest data is , while the range of years for latest data is for all countries. For countries with missing data, backward or forward forecasts were calculated based on available data points using simple linear regression. With only one data point (2005) available for Tajikistan, the value for the earliest year was imputed using the average annual growth rate in Central Asia (i.e., Kazakhstan, the Kyrgyz Republic, Turkmenistan, and Uzbekistan). Sources: World Bank. PovCalNet (accessed 19 April 2010) for poverty data; Food and Agriculture Organization (FAO). Food Security Statistics (accessed 8 March 2010) for undernourishment data; Statistics Division of the United Nations Economic and Social Commission for Asia and the Pacific (UNESCAP) for underweight children under age 5 years data for all countries except Papua New Guinea, for which 2005 data point is taken from the United Nations Children's Fund (UNICEF, 2009) and 1996 and 2006 data points are taken from ADB. Statistical Database System Online (accessed 19 April 2010); and United Nations Population Division. World Population Prospects (accessed 19 April 2010) for population data used as country weights.

11 4 ADB Sustainable Development Working Paper Series No. 11 Table 2: Absolute Numbers of Poor, Undernourished Population, and Underweight Children (million) ($1.00 PPP per day) Poor ($1.25 PPP per day) Number of Undernourished Persons b Region or Subregion Number of Underweight Children Under Age 5 Years c,d Earliest Latest ( ) ( ) Asia and the Pacific , East Asia Southeast Asia South Asia Central and West Asia Pacific a PPP = purchasing power parity. a The Pacific includes Papua New Guinea only, owing to scant data availability. b For prevalence and number of undernourished persons, East Asia s regional average includes the Democratic People's Republic of Korea and the Republic of Korea, while Southeast Asia s regional average includes Myanmar. c Country weights for proportion and number of underweight children under age 5 years were calculated by the Asian Development Bank (ADB) staff using 2005 population of children age 0 4 years from the 2008 Revision of United Nations World Population Prospects. d For proportion and number of underweight children under age 5 years, the range of years for the earliest data is , while the range of years for latest data is for all countries. For countries with missing data, backward or forward forecasts were calculated based on available data points using simple linear regression. With only one data point (2005) available for Tajikistan, the value for the earliest year was imputed using the average annual growth rate in Central Asia (i.e., Kazakhstan, the Kyrgyz Republic, Turkmenistan, and Uzbekistan). Sources: World Bank. PovCalNet (accessed 19 April 2010) for poverty data; Food and Agriculture Organization (FAO). Food Security Statistics (accessed 8 March 2010) for undernourishment data; Statistics Division of the United Nations Economic and Social Commission for Asia and the Pacific (UNESCAP) for underweight children under age 5 years data for all countries except Papua New Guinea, for which 2005 data point is taken from the United Nations Children's Fund (UNICEF, 2009) and 1996 and 2006 data points are taken from ADB. Statistical Database System Online (accessed 19 April 2010); and United Nations Population Division. World Population Prospects (accessed 19 April 2010) for population data used as country weights. The incidence of poverty for the region dropped from 52% to 27%, using the updated $1.25-a-day measure (2005 purchasing power parity or PPP). 3 Yet the incidence of undernourished population fell from 20% in to only 16% in (Table 1). Further, the extent of child undernourishment denoted by the prevalence of underweight children under age 5 years fell only from 40% to 31% in a similar period (although not the same period due to scant data, estimates based on ranges within and have been reported). 4 As far as the total number of the poor and hungry are concerned, the statistics are similar. While the number of poor has fallen dramatically from about 1.4 billion in 1990 to about 900 million in 2005 the number of undernourished has stayed roughly the same, declining only from 586 million to 566 million from to Similarly, the number of underweight children has decreased only from 127 million to 98 million. 3 The reduction in poverty, using the earlier $1-a-day measure (1993 PPP), is from 36% in 1990 to 16% in ADB UNDP UNESCAP (2007) estimated a drop for the region in the incidence of underweight children from 35% to 28% between 1990 and 2005, which shows a much smaller reduction.

12 Approaches to Combat Hunger in Asia and the Pacific Location of Hunger Hotspots by Subregion and Country Trends in poverty and hunger in Asia and the Pacific subregions 5 indicate that all have (i) experienced a decrease in the incidence of poverty (except for Central and West Asia); (ii) experienced a reduction in the incidence of hunger by proportion of undernourished population (except for the Pacific); and (iii) seen a decline in the incidence of underweight children under age 5 years, although this reduction has been less dramatic. In the case of Central and West Asia, the incidence of poverty increased, while incidence of undernourishment in the general population and of underweight children declined perhaps the result of stronger state involvement in basic services and human development. 6 For the Pacific, the poverty incidence fell, but general undernourishment marginally increased. Both of these subregions demonstrate that poverty and hunger do not necessarily go hand-in-hand. In absolute numbers, while East Asia and Southeast Asia have seen significant reductions in poverty, South Asia and Central and West Asia have seen increased numbers of poor. Further, the reduction in the number of undernourished persons has not fallen much, even in East Asia and Southeast Asia, while it has increased in South Asia and the Pacific, and declined only minimally in Central and West Asia. In the case of child hunger, the trend is similar to that of general hunger. The reduction in terms of proportion and number between the earliest figures (early 1990s) and the latest (around 2005) show a much less pronounced decline as compared to poverty. The region s divergent performance on poverty and hunger suggests that policies that attempt to reduce poverty alone may not reduce hunger. Thus, special measures are called for to address hunger. Focus is also needed on South Asia, which accounted for almost 60% of the region s hungry, 65% of the extreme poor, and 81% of underweight children. Further, South Asia is witnessing an increase in the absolute number of both poor and undernourished. Other countries where undernourishment in the general population persist (i.e., with incidence exceeding 15% by the latest count) are the Central and East Asian nations of Armenia, Mongolia, and Tajikistan; and the Southeast Asian nations of Cambodia, Indonesia, the Lao People's Democratic Republic (Lao PDR), the Philippines, Thailand, and Timor-Leste (Figure 1). By incidence of underweight children, countries other than those in South Asia that need to be monitored (i.e., having incidence exceeding 15%) are the Southeast Asian nations of Cambodia, the Lao PDR, Indonesia, the Philippines, Timor-Leste, and Viet Nam; Tajikistan in Central Asia; and Papua New Guinea in the Pacific. 5 The list of countries in these subregions is provided in Appendix 3. 6 The discrepancy could also be due to the difference in the data range endpoints for poverty and hunger.

13 6 ADB Sustainable Development Working Paper Series No. 11 Figure 1: Poverty and Hunger in Selected Asian Countries Nepal Timor-Leste Bangladesh India Cambodia Uzbekistan Lao PDR Papua New Guinea Viet Nam Philippines Pakistan Mongolia Kyrgyz Republic Tajikistan Indonesia People's Republic of China Georgia Turkmenistan Sri Lanka Armenia Kazakhstan Malaysia Thailand Azerbaijan HCR ($1.25 PPP per day) Proportion of underweight CU5 Prevalence of undernourishment (% population) CU5 = children under age 5 years, HCR = headcount ratio, Lao PDR = Lao People's Democratic Republic, PPP = purchasing power parity. Note: The latest undernourishment data for Kazakhstan, the Kyrgyz Republic, and Malaysia are based on data from an earlier release of the Food and Agriculture Organization (FAO) food security statistics. Sources: World Bank. PovCalNet (accessed 19 April 2010) for poverty data; FAO. Food Security Statistics (accessed 8 March 2010) for undernourishment data; Statistics Division of the United Nations Economic and Social Commission for Asia and the Pacific (UNESCAP) for underweight children under age 5 years data for all countries except Papua New Guinea, which are taken from the United Nations Children's Fund (UNICEF 2009). The GHI rankings also reveal a similar picture (Appendix 2). Of 26 countries in the region with a GHI ranking, all of South Asia were in the bottom half and have seen little improvement in rankings between 1992 and Several Southeast Asian countries have also seen little progress in the GHI (i.e., less than 25% during ) including Cambodia, the

14 Approaches to Combat Hunger in Asia and the Pacific 7 Lao PDR, Myanmar, and the Philippines, all of which also figure in the bottom 50% of the rankings. Variations within countries show interesting facets of hunger in the region. For instance, despite a common pattern of higher poverty in rural areas, all countries have a higher food energy deficiency incidence in urban areas (Jha et al. 2009a). This is expected, since despite higher incomes, the urban poor have little income left for food after meeting minimum expenses for housing, transport, and health. Appendix 3 provides a profile of the evolution of poverty and calorie deprivation in some countries in the region. 3.3 The Global Food and Economic Crises and Impact on Poverty and Hunger Analyzing hunger and poverty is complicated by the fact that their data are available only after considerable delay. As such, the estimates of poverty and hunger presented in this paper relate to 2005 for poverty and for undernourishment. Estimates of child undernourishment are even more problematic, as data are less regularly available. Reliable survey-based estimates of the impact of the global food and economic crises on hunger are therefore unavailable. However, some judgments can be made. The regional slowdown caused by the global economic crisis has not generally caused a contraction in regional or individual country economic growth rates, except in a few cases. Within most countries, however, there have been sharp contractions in some sectors, such as those associated with exports, with resulting unemployment. The deceleration of growth will slow down reduction in the incidence of poverty rather than increase it. In addition, some estimates suggest that an additional 21 million people in the region who would have been raised up from absolute poverty during (i.e., if the crisis did not occur) will remain trapped in poverty (ADB UNDP UNESCAP 2010). The food crisis, on the other hand, by impacting on the price of food that forms a major part of the budget of the poor, has had a more severe impact. Food prices have been rising steadily since The FAO Food Price Index, which was 111 in 2004, peaked at 214 in June Though it declined thereafter, it began increasing again in 2009, reaching 172 in December Cereal prices have stayed much higher, and also appear to be stabilizing at much higher levels than in (Figure 2). This index was 107 in 2004, peaking at 282 in May 2008, and remaining as high as 170 in December 2009 (i.e., about 60% higher than 2004 levels) (FAO 2010). Rough estimates made suggest that the incidence of hunger and poverty has risen in most countries in the region as a result of the rise in food prices. In the Philippines, for example, a 30% increase in food prices, could have resulted in a 10% increase in poverty. In fact, cereal prices were higher by 97% in 2008 than 2006 and by about 50% in The increase in the number and incidence of undernourishment resulting from both crises was confirmed by FAO, which estimated that the total undernourished population in the region increased from 566 million in to 642 million in 2009 (FAO 2009a). This also implies an increase in the incidence of hunger from 16% to 18% in this period. 7 See ADB (2008).

15 8 ADB Sustainable Development Working Paper Series No. 11 Figure 2: Food and Cereal Price Indexes = Points Jan 2008 Mar 2008 May 2008 Jul 2008 Sep 2008 Nov 2008 Jan 2009 Mar 2009 May 2009 Jul 2009 Sep 2009 Nov 2009 Jan 2010 Food price Price index Index Cereals price Price index Index Source: Food and Agriculture Organization Future Projections Regarding future trends, there is concern whether the MDG hunger targets will be reached. As far as undernourishment of the population is concerned, in its 2009 State of Food Insecurity in the World report, FAO indicated that "[w]hile the proportion of undernourished continually declined from to , the decline was much slower than the pace needed to meet the hunger reduction target of the first Millennium Development Goal" (2009a, p. 8). This is borne out by the data and trends presented in Table 1. From Table 1, it can be inferred that the region has achieved roughly a 1.6% annual reduction (geometric) in the incidence of undernourishment in the population, which is too slow to achieve the MDGs. The main stumbling block is South Asia, where the reduction has been abysmally slow at just 0.6% per year. Moreover, the region is much further away in meeting the stricter World Food Summit target of reducing the absolute number of hungry by one half from the 1990 level of 586 million to 293 million by In fact, by , the region had reduced the population of malnourished by only 20 million, to 566 million. In the case of South Asia, the numbers of malnourished have heavily increased. With the steady increase in food prices being witnessed again since 2009, the prospects of the target being reached have further receded. As far as the MDG target on incidence of underweight children is concerned, the latest projections indicate that, given current trends, the region is unlikely to achieve it (ADB UNDP UNESCAP 2010). Asia and the Pacific has a higher proportion of underweight children than even sub-saharan Africa. In fact, Asia and the Pacific's overall performance on child malnutrition was considered in an earlier report as one of the region s greatest failures (ADB UNDP UNESCAP 2007). This problem, like malnourishment in the general population, has also been further aggravated by the increase in food prices. This calls for a more detailed discussion of hunger in this particularly vulnerable group children.

16 Approaches to Combat Hunger in Asia and the Pacific 9 4. Child Hunger Child malnutrition, as reflected in body weight, is an important indicator of hunger, and is the second of the two hunger indicators in MDG goal 1. Child malnutrition measured by body weight can also be monitored more closely than adult malnutrition and is, therefore, a better indicator of malnutrition in the population. The World Bank also considers it a singularly important indicator of deprivation. 8 In many countries, child malnutrition 9 has been linked to poverty, low levels of education, and poor access to health services. The undernourishment of children at a young age has lasting impacts, with severe consequences on their cognitive development and increases their risk of death. The foundations of adequate child nutrition are laid during the mother s pregnancy. As the World Bank notes, provision of adequate nourishment to expectant and lactating mothers is crucial to the healthy development of their offspring. Among regions, Asia and the Pacific accounts for an astonishing 68.0% of the world's underweight children. 10 As few as seven countries (i.e., Bangladesh, the People's Republic of China [PRC], India, Indonesia, Pakistan, the Philippines, and Viet Nam) account for 62.6% of these children, with India alone accounting for 39.0%. In South Asia, around 40% of children under age 5 years are underweight in Bangladesh, India, Nepal, and Pakistan, with the figure somewhat lower in Sri Lanka (29%). Regarding progress on this MDG target, a study estimated that neither Afghanistan, India, Nepal, nor Pakistan will achieve the MDG target, with Pakistan regressing (ADB UNDP UNESCAP 2010). In Southeast Asia, a somewhat better-performing subregion, the larger countries of Indonesia, the Philippines, and Viet Nam continue to face significant problems relating to child malnourishment, with prevalence rates ranging from 20% to 30%. In Cambodia, the Lao PDR, and Timor-Leste, over 35% of all children under age 5 years are underweight. Apart from Viet Nam, none of these countries will reach the MDG target, and Timor-Leste is regressing (ADB UNDP UNESCAP 2010). In East Asia, the problem of child undernourishment is less severe. The PRC more than halved the proportion of underweight children between 1990 and 2005, from 19% to 7%, and may achieve the MDG target. However, the Democratic People's Republic of Korea has a relatively high proportion of underweight children, and progress in reduction has been slow. Malnourishment in children appears to be a contributing cause of reduced infant and child survival rates (Appendix 4). The correlation rate between the incidence of underweight children under age 5 years and under age 5 years child mortality is moderately high at 0.42 for the latest period reported. 8 Child undernourishment is defined as the proportion of children age 0 59 months who fall below the median weight for the age of the National Center for Health Statistics WHO standard reference population by more than three standard deviations. In a normally distributed population, only 0.13% of children would be expected to be severely underweight. See World Bank (2007a). 9 UNESCAP (2007) also noted that the number of children under age 5 years who are underweight is an internationally recognized indicator of pubic health. 10 Some of the discussion on child hunger in this section is derived from the Global Framework for Action Ending Child Hunger and Undernutrition Initiative through the REACH partnership, in which ADB has also participated. See World Food Programme and UNICEF (2007).

17 10 ADB Sustainable Development Working Paper Series No. 11 Child malnutrition requires a multidisciplinary approach. The causes, according to a recent global study by REACH (2007) are mainly threefold: (i) inadequate access to food, including nutrients; (ii) insufficient maternal and child care; and (iii) lack of essential health services. Adverse political, economic, and cultural environments also play a role, as well as inadequate information and education about proper nutrition. Required interventions suggested by REACH that need to be scaled up are (i) local and homestead food production; (ii) nutrition education; (iii) micronutrient fortification and supplementation; (iv) provision of safety nets to vulnerable households; (v) household water treatment; (vi) improved hygienic practices, particularly hand washing with soap; (vii) parasite control (e.g., deworming and use of bed nets in malaria-endemic areas); (viii) special feeding programs; and (ix) maternal care, maternal nutrition, and breast-feeding. At the same time, continuing attention must be paid to ensure national food security; disease prevention and control (e.g., improved acute respiratory infection and diarrhea treatment, and immunization programs); provision of clean water and improved hygiene; and improved reproductive health, particularly safe motherhood and birth spacing. REACH has also suggested a strategy designed to achieve four major outcomes regarding child malnutrition and hunger: (i) increased awareness and understanding of the problem and its causes and solutions; (ii) strengthened, resourced, and monitored national programs; (iii) increased capacities on the international, national, and household levels for taking action on this issue; and (iv) increased efficiency and better monitoring and evaluation of global efforts. The strategy requires incorporating programs for hunger and malnutrition at the national and subnational strategic planning levels, devoting more resources to the problem compared to current allocations, creating better synergies and partnerships among supporting national and international agencies, stoking greater involvement of and partnership with communities, improving micro-level planning to tie up delivery agencies and outreach partners with support and funding agencies, and bettering capacities at all levels involved with planning and delivery of programs. How do these suggestions square with the implementation experience of existing child malnutrition programs? India, which faces one of the biggest challenges in this regard, has introduced some of the largest child nutrition and feeding programs in the world. These include the Integrated Child Health Services Scheme and the National Midday Meal Scheme for primary schoolchildren (Box 1). Better targeting, involving communities, and adopting a multidisciplinary approach are all aimed to help improve the effectiveness of the Integrated Child Health Services Scheme. The National Midday Meal Scheme has demonstrated benefits for schoolchildren; however, the problem of preschool-age child malnutrition remains a challenge and cannot be corrected by programs for older children. Box 1: Integrated Child Development Services and National Midday Meal Schemes in India A 2006 World Bank study by Gragnolati et al. highlighted some basic characteristics of child malnutrition in India: it is higher in (i) rural than in urban areas, (ii) female children than in male children, (iii) lower castes and tribal households than in the general population, and (iv) the poor than in the general population. About 60% was in the poorest wealth quintile. Six states (i.e., Bihar, Madhya Pradesh, Maharashtra, Orissa, Rajasthan, and Uttar Pradesh) accounted for the bulk of underweight children, and one-quarter of all districts accounted for more than half. Micronutrient deficiency was also widespread, with 75% of preschool-age children suffering from iron-deficiency anemia and 57% from vitamin A deficiency. Iodine deficiency was endemic in 85% of districts. While progress on reducing child malnutrition has occurred, it has been far slower in India than other countries experiencing similar economic growth. Clearly, economic growth alone cannot be relied upon to solve this problem.

18 Approaches to Combat Hunger in Asia and the Pacific 11 Against this background, the World Bank found weaknesses in the Integrated Child Development Services Scheme. First, although the scheme was designed to address the multidimensional causes of undernutrition, it focused predominantly on supplementary feeding and preschool education for children 4 6 years old, failing to address children under the age of 3 years (i.e., those who suffer most from malnutrition) and to counsel parents on better feeding and child care practices. More attention to disease control and prevention activities, education to improve domestic child care and feeding practices, and micronutrient supplementation are needed. These will require restoring the multidisciplinary approach of the scheme, as well as integrating activities into the health sector and the Reproductive and Child Health Program. Second, better targeting is needed toward the most vulnerable age groups (i.e., children under age 3 years and pregnant women), and to states and districts with the highest prevalence of malnutrition. Similarly, supplementary feeding activities need to be aimed to those who need it most. Third, greater community involvement is necessary to augment resources and to improve the quality of service delivery and accountability. Finally, monitoring and evaluation activities need strengthening. In 1995, the National Midday Meal Scheme was introduced but did not become a universal feeding program for primary schoolchildren until a 2003 Supreme Court ruling. Under this scheme, all children in public schools receive a cooked meal, consisting of no less that 300 calories and 8 12 grams of protein. Surveys quoted by Singh (2008) confirmed the benefits from this scheme, including increased enrollment rates, especially benefiting girls and lower castes, and reduced calorie and protein deficiency. Studying the scheme's impact in Andhra Pradesh, Singh also found that there were long-term improvements in cognitive skills. However, the scheme did not have strong long-term nutritional effects on body weight and height, except for younger children affected by drought. Sources: Gragnolati et al and Singh Causes and Correlates of Hunger 5.1 Poverty and Hunger Although most assume that poverty is the central cause of hunger, the reduction in hunger has not matched the decline in poverty in the region. Poverty and hunger are expected to be highly correlated, since an overwhelmingly large proportion of the poor s expenditure is on basic necessities like food. Poor people not only struggle to acquire adequate quantities of food, but also face food insecurity because their diet often lacks essential micronutrients. Hunger is also a central cause of poverty as well, although this causal link is less recognized. Undernourished people are less productive, while child malnutrition has severe and permanent consequences for physical and intellectual development. Undernourished people and children severely impair their future earning capacity. Thus, the effects of poverty and malnutrition can often last over several generations. Regressions, using cross-country data for for 43 countries, do not show a strong relationship between undernourishment in the general population (measured by the FAO index) and poverty both the $1.00- and the updated $1.25-a-day measures (Table 3). The regression of the logarithmic values of these variables do not yield a good fit (low R 2 values). On the other hand, the relationship between poverty and incidence of underweight children (measured by the UNICEF index) is stronger, indicating that poverty is an important cause of the incidence of underweight children in the region. Correlation between the updated $1.25-aday poverty measure and malnourishment in the population is only 0.24, while it is 0.72 using

19 12 ADB Sustainable Development Working Paper Series No. 11 the measure for underweight children. Similar correlations are observed with $1.00 poverty (0.21 and 0.67, respectively). Table 3: Regressions of Poverty and Hunger, Asian Countries ( ) Poverty Headcount Ratio (PPP $1.00 per day) (PPP $1.25 per day) Dependent Variable Coefficient P-Value R 2 Coefficient P-Value R 2 Proportion of underweight children under age 5 years Undernourished population (% of total) PPP = purchasing power parity, R 2 = coefficient of determination. Sources: World Bank. PovCalNet (accessed 4 August 2010); Asian Development Bank ( ) Key Indicators for Asia and the Pacific Series; World Development Indicators Online (accessed 4 August 2010); and Hasan et al Ahmed et al. (2007a), using survey data of 16 countries and defining the hungry as those consuming less than 2,200 calories a day, found a higher correlation (0.63) between hunger in the general population as measured by calorific deficiency and poverty. However, analyzing the disaggregated results for subjacent, medial, and ultra poverty and hungry groups shows that the correlations between poverty and hunger are relatively high and statistically significant only for the ultra poor and hungry. The correlation between the incidence of poverty (i.e., all subgroups combined) and the incidence of low nutrition quality in diets 11 are found to be weakly positive (0.43). One possible reason why the two seemingly close phenomena of poverty and hunger fail to show high correlation is that the measurement of both involves issues that have not been fully resolved. Box 2 discusses some of these data and measurement difficulties. Box 2: Data Limitations in the Computation of Hunger and Poverty Computing poverty and hunger indexes is difficult. Regarding poverty, problems include (i) differences in consumption figures computed from household-level data and those from national data (Ravallion 2003), (ii) defining an acceptable global poverty line, and (iii) the existence of measurement errors in the data (Reddy and Pogge 2005; Pogge and Reddy 2006). The new $1.25-a-day measure, based on a 2005 comprehensive international comparison project, may help estimate poverty more accurately. Challenges also exist in measuring hunger. Food and Agriculture Organization (FAO) methodology for computing hunger, for instance, has several issues. It comprises estimating (i) the national per-capita availability of calories; (ii) the distribution of available calories across households; and (iii) a minimum cutoff point as the lowest acceptable per-capita availability (intake) of calories, below which households are classified as undernourished (Svedberg 1999). However, the quality of data from different regions varies greatly, and while the survey techniques used are designed to minimize error, the sample sizes are small, and sampling is not random. Finally, the estimation of the cut-off point involves a complex procedure and 11 This is a measure of nutritional quality, and the study considers one to have a low-quality diet if he or she consumes food from less than five of the seven essential food groups.

20 Approaches to Combat Hunger in Asia and the Pacific 13 relies on a wide range of difficult-to-measure variables, for which FAO uses a subjective best judgment. According to Svedberg, the FAO methodology is simplistic; provides a supply-side explanation of undernourishment, shedding little light on demand-side factors; and cannot properly identify undernourished individuals and households. Svedberg's alternative method anthropometrics has advantages in that it is based on data collected for individuals; allows estimation of intra-household allocation of nutrients by age and gender, which is useful for targeting households and individuals in intervention schemes; and is simple and accurate. However, data requirements are quite demanding, relying on family health surveys rather than household surveys. These data difficulties may partly explain why poverty reduction is not observed to translate into hunger reduction. In particular, the relationship between prices for food and non-food items in a price index used to measure poverty influences how poverty translates into hunger. For example, if non-food items are more expensive and are given greater weight in the index, then the index will be higher, causing more people to be measured as poor, even though their nutritional needs are met. Conversely, if nonfood items are relatively cheaper (as is currently happening with prices falling due to better technologies), then fewer people may be classified as poor, even though they are nutritionally deficient. Such factors may weaken the relationship between hunger and poverty. Sources: Pogge and Reddy 2006, Reddy and Pogge 2005, and Svedberg Economic Growth Economic growth has been an important factor in the reduction of poverty in Asia and the Pacific. However, the impact of economic growth on hunger in general has been much less pronounced, suggesting (i) that the quality of economic growth and particularly its greater inclusiveness is important, and (ii) that more instruments than merely striving for higher economic growth are necessary to combat hunger in the region. Cross-country regressions, using per capita gross domestic product (GDP) as an explanatory variable on the two MDG indicators of hunger in 43 countries and on the two poverty indicators (i.e., the $1.00-a-day and the updated $1.25-a day markers) reveal that the association of growth to poverty reduction is substantially stronger than to reduction of hunger. Thus, the regressions (Figure 3) reveal an elasticity of 2.1 between per capita GDP and poverty measured by $1.00-a-day for the countries studied, similar to that obtained in ADB's Key Indicators (2004a). On the other hand, the responsiveness of the updated $1.25-a-day poverty measure is less strong, with an elasticity of 1.7. This bears out with the now better-estimated and larger magnitudes of poverty that are revealed to be present despite rapid economic growth UNDP (2009), however, found a stronger association between hunger and national poverty data.

21 14 ADB Sustainable Development Working Paper Series No. 11 Figure 3: Regression of the Change in Poverty Headcount Ratio ($1.00 per day) and Gross Domestic Product per Capita Levels Figure 4: Regression of the Change in Poverty Headcount Ratio ($1.25 per day) and Gross Domestic Product Per Capita Levels ln_hcr1ppp ln_hcr125ppp ln_gdpcap ln_gdpcap Note: Trend line represents fitted values. ln_gdpcap = Natural logarithm of gross domestic per capita ln_hcr1ppp = Natural logarithm of poverty headcount ratio ($1.00 per day) ln_hcr125ppp = Natural logarithm of poverty headcount ratio ($1.25 per day) Sources: World Bank. PovCalNet (accessed 4 August 2010); ADB ( ) Key Indicators for Asia and the Pacific Series; World Development Indicators Online (accessed 4 August 2010) On the other hand, a 1.0% growth in per capita income leads to only a 0.5% reduction in the incidence of underweight children below age 5 years and an even lower 0.2% reduction in the incidence of undernourishment in the overall population (Figure 5). Using subregional dummies to compare structural subregional differences to a base subregion, East Asia, reveals that growth has even less impact on underweight children in South Asia than in East Asia, while it has more impact in Central and West Asia. As far as overall undernourishment of the general population is concerned, the impact of growth on undernourishment is stronger. These are important conclusions for policy. Some have also argued for the existence of a reverse causation from hunger to economic growth as well as for the existence of a growth nutrition trap (FAO et al. 2005a and 2005b). Hunger adversely affects health and productivity, lowers output, and increases health-related expenses that reduce economic growth, further increasing hunger. FAO (2005b) indicated that the present discounted value of losses, due to undernourishment from a cohort of undernourished children over a lifetime, adds up to around $500 billion $1 trillion, or roughly 5% 10% of GDP of the developing world. Moreover, the developing world annually spends about $30 billion on health problems associated with hunger; this amount is more than one third of total official development assistance. Thus, the hunger low-productivity extreme poverty nexus acts as a trap for the poor and hungry Jha et al. (2009a) provided evidence of such a poverty nutrition trap for calories and micronutrients in India.

22 Approaches to Combat Hunger in Asia and the Pacific 15 Figure 5: Regression of the Change in the Proportion of Underweight Children under Age 5 Years and Gross Domestic Product per Capita Levels ln_underwt Figure 6: Regression of the Change in the Proportion of Population below Minimum Level of Dietary Energy Consumption and Gross Domestic Product per Capita Levels ln_undernour ln_gdpcap ln_gdpcap Note: Trend line represents fitted values. ln_gdpcap = Natural logarithm of gross domestic per capita ln_underwt5 = Natural logarithm of proportion of underweight children under age 5 years ln_undernour = Natural logarithm of proportion of population below minimum level of dietary energy consumption Sources: World Bank. PovCalNet (accessed 4 August 2010); ADB ( ) Key Indicators for Asia and the Pacific Series; World Development Indicators Online (accessed 4 August 2010) Arcand (2001) studied this relationship using two measures of nutritional status defined by FAO: (i) prevalence of food inadequacy, which is seen to have a negative effect on economic growth; and (ii) dietary energy supply per capita, which has a positive effect. The study found a statistically significant and quantitatively important impact of nutrition on growth. It also found nonlinearities between nutrition and economic growth, 14 suggesting that for a country with an already high GDP per capita, increasing nutrition would have little impact on economic growth. In other words, as countries improve their nutritional status, the gains to economic growth from this will eventually be exhausted. In addition to the initial GDP per capita, the initial nutritional status of the population is also important. The findings revealed that if initial consumption is below a certain critical level, then undernourishment will adversely affect productivity, which, in turn, will keep per capita GDP low, thus depriving people of the economic means to improve their nutritional profile and to climb out of poverty. Arcand found considerable support for the nutrition poverty growth trap. His results also suggest that average GDP per capita growth would have been higher in developing countries and more equally distributed if minimum calorific norms were attained. 14 The threshold level of the dietary energy supply per capita at which its marginal impact on growth begins to become negative is equal to 3,066 calories per day, which is quite high. For the prevalence of food inadequacy, experimentation revealed a cubic specification, although this was not found to be universally statistically significant in the full range of tests.